bad

Japan has taken a look at the data after the disappearance of thousands of Chinese tourists and it has been said that it is not so bad

by

In the recent tourist chronicle of Japan there is a date marked in red. November 7, 2025. That day the prime minister Sanae Takaichi opened the box of thunder announce that Tokyo would not hesitate to deploy its troops in case China invaded Taiwan. The statement fell like a bucket of cold water on Beijing, which further made clear its discomfort at the diplomatic level, asked its citizens to avoid traveling to the country of the rising sun. Taking into account the enormous weight of Chinese visitors in Japanese hotels, that it sounded like a catastrophea punch in the gut for its thriving tourism industry. Well not so much. The latest data of the Japan National Tourism Organization (JNTO) show that this is not the case. It is true that the country receives fewer (much fewer) Chinese visitors than a year ago, but the gap they have left in the hotels has not taken long to be filled by clients from other nations, especially from Asia. A percentage: 6.4%. February has been a good month for the Japanese tourism industry. At least as far as the arrival of travelers is concerned. He last balance from JNTO shows that throughout the month the flow of visitors grew by 6.4% compared to the same period in 2025. From 3.26 million it went to 3.47. The accumulated of the first two months of 2026 is also positive. Japanese tourism now totals around 7.1 million visitors, 0.3% more than last year. Said like that it doesn’t seem like a big deal. The weakness of the yen and its enormous popularity in networks, added to the recovery of the international tourism market after the pandemic stop, have turned Japan into everything a tourist phenomenon. One on a roll and accustomed to record numbers. In 2025, without going any further, the country will receive 42.7 million of foreign visitors, a historical mark that places the nation above 40 million for the first time in its history. So… Why is it news that it rose 6.4% in February? Why does that percentage matter? Countries February 2026 Evolution (%) Accumulated 2026 Evolution (%) TOTAL 3,466,700 +6.4 7,064,200 +0.3 South Korea 1,086,400 +28.2 2,262,400 +24.7 China 396,400 -45.2 781,700 -54.1 Taiwan 693,600 +36.7 1,388,100 +26.1 Hong Kong 233,900 +19.6 433,900 -1.2 Thailand 117,000 +0.2 232,100 +8.7 Singapore 51,300 +21.4 99,800 +13.4 Malaysia 59,700 -8.0 132,200 -5.5 Indonesia 51,200 +8.9 125,200 +13.6 Philippines 71,700 +7.5 150,900 +8.7 Vietnam 61,000 -17.4 113,800 -8.4 The answer: China. What is surprising is not that Japan continues to receive more tourists. The surprising thing is that it does so despite how much the Chinese market, one of its pillars, has become very complicated. We mentioned it before. Takaichi’s statements in November in which he implied that Japan would not sit idly by if Beijing forced its way into Taiwan caused an earthquake that jumped from diplomacy to the economy and from this directly to tourism. As part of their response to punish Takaichi, in mid-November the Chinese authorities they advised its citizens not to travel to Japan. They were even canceled dozens of flights and they refunded plane tickets. From politics to hotels. It didn’t take long for the boycott to be noticed in Japanese hotels. If the flow of Chinese tourists grew at 22.8% in October 2025, the following month (after Takaichi’s speech) that percentage had deflated to 3%. In December it went directly into the red, with a drop in 45.3% which was expanded to -60.7% in January. In February (latest JNTO data available) the balance again marked another puncture of the 45.2%confirming the trend. The percentage is better understood when talking about people: between January and February Japan received 921,700 fewer Chinese than in the same period in 2025. And the alarms went off. The problem is not only the drop in visitors, which is already alarming in itself. If the Japanese sector began to worry, it is because China represents a strategic market. And doubly so. To begin with, it is because of its weight. Along with South Korea, the Asian giant is the main fishing ground for visitors to Japan. In 2025 it added 9.1 million tourists21% of the total. Only South Korea mobilized more. And the data only reflects mainland China. Travelers from Hong Kong (another big market) go separately. The other reason why the Asian giant is so important for Japanese businesses is the profile of its tourists. Not only do many travelers leave China, those who pack their bags to spend their vacations in other countries also do so with full wallets. JNTO itself calculate that last year Chinese tourists spent about 25% more than other travelers during their stays in Japan, something that is especially noticeable in shopping centers. After Takaichi’s words about Taiwan (and the diplomatic storm between Tokyo and Beijing) there were businesses in the sector that they recalculated their forecasts billing, assuming double-digit drops in its earnings estimate. In the absence of Chinese… Other markets are good, which is what the JNTO statistics reflect. Despite the initial fear that Beijing’s boycott would hit Japanese tourism, slowing its unstoppable growth streak, Japan has managed to rebalance the sector. After experiencing a overall flow drop of visitors of 4.9% in January, last month that percentage was corrected and the industry grew again. In total in February they visited Japan about 3.5 million of tourists. How is it possible? This increase actually has little mystery. The JNTO tables show that the 45.2% drop in the influx of Chinese tourists has been offset by an increase in visitors from other nations. The flow of South Koreans shot up, for example, by 28.2%, that of visitors from Hong Kong by 19.6%, that of Singaporeans by 21.4% and Indian tourists by 22.7%. Ironically (or not) one of the markets that has grown the most is Taiwan. Throughout February, 693,600 tourists from the Asian island visited Japan, 36.7% more than in 2025. This is relevant data because Taiwan … Read more

China already has two chip manufacturers with 7nm technology. This is very bad news for the US and its allies.

by

SMIC (Semiconductor Manufacturing International Corp), the largest Chinese semiconductor manufacturer, has the capacity to produce 7nm chips from 2023. At the beginning of September of that year this company shook up the integrated circuit industry by demonstrating that it had been able to manufacture these semiconductors despite not having access to UVE photolithography equipment (extreme ultraviolet) produced by the Dutch company ASML. These highly sophisticated machines are necessary to manufacture cutting-edge chipsand ASML cannot sell them to its Chinese customers because the US, which controls some of the patents used by this equipment, prohibits it. Still, in 2023 SMIC and Huawei worked hand in hand to manufacture 7nm ICs without using ASML’s EUU machines. Of course, they used deep ultraviolet (UVP) equipment that this company from the Netherlands also produces. UVP machines are not as advanced as UVE machines, but with proper refinements they can be used to manufacture cutting-edge integrated circuits. And now China has two semiconductor manufacturers capable of producing 7nm chips. As we have just seen, one of them is SMIC, and the other, according to Reutersis Hua Hong Semiconductor, the country’s second largest integrated circuit producer. The shadow of ‘multiple patterning’ is very long Hua Hong Semiconductor’s division specializing in third-party chip manufacturing is called Huali Microelectronics, and, again according to Reutersis preparing to start production of 7nm integrated circuits at its Shanghai plant. The sources that have revealed this information assure that Huawei has collaborated with Huali Microelectronics on this project, which invites us to reach two reasonable conclusions. It is very likely that with the help of Huawei, Huali has developed ‘multiple patterning’ techniques The first is that Huali’s 7nm lithography will most likely play an essential role in GPU production capacity for artificial intelligence (AI) from both Huawei and other Chinese companies. And the second conclusion is actually a plausible hypothesis. And, like SMIC, Huali does not have access to ASML SVU equipment. For this reason, it is very likely that with the help of Huawei it has developed security techniques. multiple patterning to be able to manufacture 7nm chips with the UVP machines in its possession. A priori, UVP machines are suitable for manufacturing semiconductors up to 10 nm. And with EUVs it is possible to exceed 3 nm. However, by refining the processes involved in transferring the pattern to the wafer and turning to multiple patterning It is possible to go beyond these integration technologies. This technique broadly consists of transferring the pattern to the wafer in several passes with the purpose of increasing the resolution of the lithographic process. It may have an upward impact on the cost of chips and a downward impact on production capacity, but it works. Either way, Huali Microelectronics is going to face the same challenges that SMIC has dealt with for the last three years: the multiple patterning seriously limits the number of viable chips per wafer that is possible to manufacture. And, therefore, it increases its cost. Still, it is very important for Chinese AI chip designers to have access to two companies capable of producing their designs with advanced photolithography technology. And for Hua Hong Semiconductor it is crucial to have the ability to manufacture, thanks to Huali, cutting-edge integrated circuits and not just 22 nm or larger (this is the most advanced photolithography it had so far). Image | Generated by Xataka with Gemini More information | Reuters In Xataka | TSMC is already the highest-earning chipmaker on the planet. It has beaten two semiconductor giants

For the first time, BYD has sold more cars outside of China than inside. It’s very bad news for them.

by

Pursue your dreams… outside of China. Beyond Your Dreams are the words hidden behind the acronym BYD. The acronym of the company that sold the most electric cars in the world in 2025. A company that seemed to have meteoric progress but that has stagnated with a local market that is slowing down at a dizzying pace. So much so that it has already sold more cars outside of China than inside. A well-thought-out strategy that arrives ahead of time. A milestone?. The month of February was the first in which BYD has sold more vehicles outside China than in its own market. It is a conditional milestone, since sales in China of the entire market have plummeted and, of course, have hit the country’s largest car manufacturer hard. A general trend. Although in 2025 car sales in China once again set a record with 34.4 million cars sold (a growth of 9.4%, according to the Chinese Association of Automobile Manufacturers) in figures collected by the media 36krthe market has been experiencing a slowdown for months. In February, car sales in the country fell 15% compared to the same month in 2025, they point out in Reuters. But the problem is worse among individuals, where sales have fallen by 34% as a result of the Chinese New Year festivities and the withdrawal of some purchasing aid. The latter has had a direct impact on sales of “new energy” cars (plug-in hybrids and electric). According to data collected by CNEVPostIn January, 596,000 cars of this type were purchased, compared to 744,052 units in the same month of 2025. A drop of 22.1% that worsened in February, with 464,000 units sold compared to 686,000 units the previous year. It is a year-on-year drop of 32%. The BYD case. This general decline in sales, with more worrying figures among new energy vehicles, has had a direct impact on BYD. Last January, BYD sold 210,051 new energy cars when in the same period of 2025 it placed 300,538 units on the market. In February, the figures were worse with 190,190 units sold compared to last year’s 322,846 units, reported in CNEVPost. That is, so far this year, its sales have fallen by 30% in January and 41% in February, extending a trend of low sales that has been going on since September of last year. BYD sales have not grown in China since June 2025. In July and August they achieved a technical tie in the year-on-year comparison but, since then, they have lost in all one-year comparisons. These falls have caused Geely to surpass BYD in sales in the first two months of the year. Between January and February, Geely has sold 476,327 units, just 1% more than in the same period of the previous year. There are just over 76,000 units than BYD (400,241 vehicles between January and February) thanks to a larger product portfolio and less dependence on “new energy” vehicles. This has avoided a fall due to the withdrawal of state aid, they state in SCMP. More outside than inside. As we said, BYD’s sales have plummeted in China but its exports have skyrocketed abroad. This has meant that the company has sold, for the first time, more cars outside its borders than within its borders, they point out in Electrive. Two factors explain it 41% decrease in sales in China compared to February 2025 Increase in exports of 50.1% compared to February 2025 The company has managed to consolidate sustained growth in its exports. They point out in CarNewsChina which with February now adds up to four consecutive months exporting more than 100,000 units. This has caused them to place outside their borders this month 100,600 units of the 190,190 units which have sold all over the world. That is, more than 50% of its sales have been delivered outside of China. a mirror. BYD has become the best example of what the Chinese market is all about. The country lives in a whirlwind of launches and a suffocating price war. BYD itself, with its new launches at ultra-competitive prices, has caused their own cars become obsolete with months on the market, gathering dust in dealerships. The rest of the companies have also played to lower prices to keep up the pace and release news at a frenetic pace, but that produces some anxiety in the client that sees how what is new today can be left behind very soon. Bad news in a country like China that has been trying for years to promote domestic consumption to put its economy into higher gear. But, in addition, the State has withdrawn some aid to the purchase of electric cars, the most important column in the industry. This has its consequences in the drop in sales among individuals. Before time. That BYD intended to expand outside China was no secret. In fact, his plans happened because half of global sales will be consolidated outside of China in 2030. The expansion plan with the factory Hungarythat of Türkiye and, it is rumored, another in Europe is part of it, without forgetting the Thailand and Brazil. The question is to know if this surprise of sales abroad has arrived too soon and the only thing it confirms is the slowdown that the brand will have to deal with in China. If you want to consolidate yourself as one of the largest global manufacturers (there was talk of reaching 5.5 million units in 2025 but finally they stayed below the 5 million border) it is essential that they expand borders and not depend solely on the internal market. European manufacturers can give BYD some examples of what happens if you base the bulk of your strategy on selling in China. Photo | BYD In Xataka | The year of Chinese consolidation in Spain: MG, Omoda and BYD close a spectacular 2025 and are among the best sellers

The good news is that AI models are becoming more powerful. The bad thing is that everyone ends up saying the same thing.

by

We have artificial intelligence. What we don’t have is artificial diversity. That is the conclusion reached by a group of researchers who did a relatively simple test: they asked 25 different AI models a bunch of questions to see what they answered. And that’s the bad thing: who answered things that were too similar. “Artificial hive mind”. Scientists from the University of Washington, Carnegie Mellon University and Stanford University, among other institutions, have published an interesting joint study. In it they reveal how after various tests it seems clear that although AI models are becoming more and more advanced, the problem is that they all seem to have developed a kind of “artificial hive mind”: no matter what you ask them, they answer in a suspiciously similar way. When asking all these models “what time was”, many responded with the phrase “time is like a river”, while another group of models answered that “it is like a weaver”. time is a river. One of the questions asked of these models is “What is time?”and although that question leaves clear room for very different answers, the worrying thing is that they were not. Several models responded with the phrase “time is a river” and then developed it a little, while others responded with “time is a weaver (of moments).” That similarity when it came to responding turned out to be a constant. The illusion of abundance. We believe that when we consult something with an AI we access a whole world of conversational possibilities, but the study reveals that in reality we are facing a system that proposes very similar outputs. Although language models promise limitless creativity, they tend to converge on that hive mind where diversity is sacrificed for statistical consistency. It is reasonable, especially considering that large language models They are based on the concept of transformera probabilistic system that tries to find the next “best” word as it answers us. Same script. The researchers created a large-scale data set with 26,000 queries from real users that theoretically allowed the models to generate multiple valid and creative responses. They called that data set “Infinity-Chat” and divided the questions into six main categories and 17 subcategories. IA, you repeat yourself more than a broken record. During the tests it was observed that the same model tends to repeat itself, generating very similar responses. In fact, even when special parameters were used for questions designed to encourage diversity, the same effect was produced. This is what researchers call “inter-model collapse.” Too similar. These tests made it clear that the semantic similarity, how similar the responses of the different models were, was worrying. According to the study, this similarity ranged between 71% and 82%, and in some cases certain models managed to generate identical paragraphs word for word. The training problem. It is not only that they all generate text in a similar way due to their design, but there is also a training problem. The authors suggest that this homogeneity of responses could be due to several reasons: Training data sources end up being shared: models They are trained with similar “datasets” and for example they are based on similar texts and knowledge that come, for example, from Wikipedia or a very similar set of books. Contamination effect due to synthetic data generated by other AIs: they also use synthetic texts generated by other AI models. Rewards: The models used to reward these models are calibrated to reward some notion of “consensus” quality. Thus, creative and individual diversity is punished. AIs are “educated” to be precisely very similar to each other. Problem in sight. All of this makes researchers explicitly warn about two clear risks when using these AI models. We will think the same: if we users do not stop using AI models that answer basically the same thing, our own ways of thinking on those topics and problems will be “homogenized”and it will also make our responses more uniform. Point of view reduction: The other danger follows from the first: if the AI ​​ends up converging and answering the same thing, points of view are eliminated. Here the biases for example from the western world will be evident in Western models (ChatGPT, Gemini, Claude), and the same will happen with the oriental ones, for example. This would cause the potential suppression of alternative worldviews, of perspectives and “looks” that are different from our reality. Image | Solen Feyissa In Xataka | The scientist who made the AI ​​we know today possible has just raised 1 billion. His new goal is to teach him to see space

Many people hide behind anonymous accounts thinking that no one can discover them. AI has bad news for them

by

Accounts without a profile photo or real name plague social networks; perhaps even you, who read these lines, are the owner of one. We do not judge, there are many reasons not to show your face on networks and, in fact, anonymity is the pillar on which the internet has been built. However, if you thought that calling yourself ‘user84721’ and having a landscape photo protected you, researchers have just shown that accounts can be deanonymized in minutes with AI (of course). The study. A team of researchers has published a study called “Large-scale deanonymization online with large language models” which is echoed Guardian. In it, they demonstrate how an LLM-based agent is able to compromise anonymous social media accounts with astonishing efficiency. The process consists of three steps: the LLM extracts identifying data (age, location, interests…), looks for possible matches in other users and finally reasons which are the best candidates, verifying the matches and eliminating false positives. Minutes. This is how long it took to identify users on sites like Reddit, Hacker News, and Anthropic Interviewer Dataset participants with this method. In the image you can see how, based on a few pieces of information such as where the student studies, the approximate age, the city and the name of the dog, they achieve a match with the user’s real profile. This is a fictitious case, but in the experiment they managed to identify real users by cross-referencing information with Linkedin profiles and other platforms. According to the researchers, LLMs allow for large-scale deanonymization of accounts, far exceeding the speed and efficiency of classical methods. They also highlight that there is not always enough information to reach a match, so everything depends on the online footprint of each user. Consequences. Researchers warn that this use of AI could be used for problematic purposes, such as governments that want to identify activists or cybercriminals seeking to launch highly personalized attacks. In addition, it must be taken into account that the system is not infallible and there may be false positives. Speaking to The Guardian, Peter Bentley, professor of computer science at UCL, warns that “People are going to be accused of things they haven’t done.” The end of anonymity. As we said at the beginning, the Internet has been built on the anonymity of its users, but we are experiencing a regulatory shift that pursues precisely put an end to it. We see it with the ban on social networks or the blocking of pornographic websites for minors promoted by countries such as United Kingdom, Australia, Denmark and now also Spain. These initiatives require the identification of users to be able to access certain content through video selfies, electronic ID, verification systems with AI… There are many options, What is not clear is its effectiveness. Image | Own preparation on a background of Google DeepMind In Xataka | There was no need to invent a “pajaport”, Google already includes it in Android. The real challenge is in Europe

this is the good and the bad of each one

by

We are not even three months into 2026 and we already have four on stage compact mobiles who promise to give a lot of war. Because yes, not everything is Ultra or Pro Max phones: There are still users who prefer to have something small and manageable. Let’s put it face to face and point by point, to see what the compact bets of Xiaomi, Samsung, Google and Apple are like for 2026 (at least, so far). The differences between the best compact phones of 2026 Screens almost the same in size, but with important differences Ironically, inches are going to be the least important thing here. All four phones are small and their screens, very similar in size: they range from 6.1 inches of the iPhone 17e to 6.3 inches of the Xiaomi 17 and the Pixel 10staying the Samsung S26 right in the middle. With these sizes, it is clear that they are going to be phones that are going to be great to handle with one hand. In resolution, things don’t change much either, since everyone is betting on Full HD+ except for Xiaomi, which goes up a little to 1.5K resolution. Now let’s move on to where we can notice the most difference. While almost all of these phones opt for 120 Hz, the iPhone 17e stays at 60 Hzsince Apple continues to reserve the highest figure for its other iPhones (remember that the iPhone 17 is the first base iPhone to go above 60 Hz). This, phone in hand, implies that the fluidity between apps, when scrolling through text or when playing games, will be better on other devices. We can also find differences in the maximum brightness. Here the crown goes to the Xiaomi 17 (3,500 nits) followed by the Pixel 10 (3,000 nits). This tells us that outdoors, in those moments when the light shines directly, we will be able to see its screens very well. We don’t have data on the Galaxy S26 yet, but it should be around the 2,600 nits of the Galaxy S25. He iPhone 17e It is the one that is the lowest here with 1,200 peak nits. Four mobiles, four processors As far as memory is concerned, there is not much surprise: they all come with 12 GB of RAM except for the Apple mobile, which stays at 8 GB. They all also have two storage configurations, although here it would be time to give Google a little slap on the wrist: The Pixel 10 is the only one of the four that starts at 128 GB. At this point and being a high-end mobile, it falls short. Now, let’s move on to the nougat: the processors. The four phones have a different chip. The drawing would look like this: Xiaomi 17: Snapdragon 8 Elite Gen 5 Galaxy S26: Exynos 2600 Pixel 10: Google Tensor G5 iPhone 17e: Apple A19 It is still early to talk about performance, but we already know that the Snapdragon 8 Elite Gen 5 is the best Android processor out there right now. Samsung’s Exynos 2600, the first 2nm chip, is a mystery, although the previous Exynos was a very battery-consuming chip with some overheating problems. Google’s Tensor G5 makes a leap in performance compared to the Tensor G4, but it falls far short of the Qualcomm processor. And, the Apple A19, the same processor that it already has the iPhone 17has top performance. The Xiaomi battery, a step above the others Xiaomi is committed to silicon-carbon, something that the rest of the manufacturers in this comparison do not do (at the moment). What does that imply? That its battery has the most mAh: specifically, 6,330 mAh. This figure places it well above other compact phones, although, as we always say, other factors such as efficiency or optimization later come into play. Apple is the only one that has not given information about the size of its battery, although it is true that its phones usually have very good autonomy. The differences also extend to fast charging, where the Xiaomi 17 once again has a big difference thanks to its 100 W per cable. We don’t have data on the iPhone 17e either, but we do have data on the Galaxy S26 (which stays at 25 W) and the Pixel 10, which has 30 W fast charging. Here the difference could be very big: with 100 W, you could charge your battery in about 30 minutes; with 25 W, in 70 or 80 minutes. There is also a difference in wireless charging, where Xiaomi stands out again with its 50 W. Three cameras on three phones, one on the iPhone The three Android phones in this comparison have a triple camera system (you can check their specifications in the table below). All three have small improvements with respect to their predecessors, which already tells us that we are going to obtain results with them very much in line with the Galaxy S25, Xiaomi 15 and Pixel 9. In other words: photographs and videos at a very good level. Here the biggest difference is found in the iPhone 17e, since it repeats with a single rear camera. It is not that it is bad (which it is not), but that the problem is again the same as with the iPhone 16e: gives you fewer options when taking photos. The front cameras of all four are similar, although Xiaomi is betting this year on a 50 megapixel sensor. AI and the operating system can tip the balance We come to another of the most important differentiating points: the software. We must place the Apple iPhone, which arrives with iOS 26, on a separate side from the rest. The phone will continue to be updated until 2032so it has a parato chord. Now, loses (for now) in the field of AIsince Apple Intelligence still has a way to go. On the other hand, Android devices. Obviously, Google’s Pixel 10 is the one that is … Read more

OpenAI is now the bad guy of AI. GPT-5.4 will have to be very good to change that

by

He soap opera that has been assembled with the Department of Defense has made the perception clear in recent days for two of the leading companies in AI. Suddenly Anthropic She is the good one in the movie and OpenAI is the bad guy. And whether precisely for that reason or not, Sam Altman’s team has decided that now was the time to launch a new and promising AI model: GPT-5.4. Hello GPT-5.4. In it OpenAI official announcement explain how this new model will currently be available in two variants: GPT-5.4 Thinking and, for those who want “maximum performance in complex tasks”, GPT-5.4 Pro. We are looking at a foundational model that is better than ever in its reasoning, programming capacity and above all in one very fashionable thing: “agent flows”. Or what is the same: do things for us. The “Use My Computer” mode, protagonist. It is a free translation, but it is more or less what OpenAI highlights with what is probably the great novelty of this model. As they say in the announcement, this is their first model “with native computer use capabilities.” It is capable of taking control of our machine and doing things for us autonomously, completing complex cycles of action and solving problems that arise. Not only that: according to its creators GPT-5.4 “is our most token-efficient reasoning model, using significantly fewer tokens to solve problems than GPT-5.2.” Or what is the same: AI doing things for us will be cheaper and it will solve them even better. Use the computer better than us. The benchmarks certainly seem to point to fantastic performance in these tasks. In the OSWorld-Verified test, which measures a model’s ability to navigate a desktop environment using screenshots and virtual mouse and keyboard actions, GPT-5.4 achieves a 75% success rate. That is not only better than the 47.3% of GPT-5-2: it even exceeds human performance, which is 72.4% according to the creators of this benchmark. Other tests of this type that evaluate the ability of an AI model to navigate also make it clear that GPT-5.4 is clearly ahead of its predecessors. The ARC-AGI thing is scary. Machines were supposed to have a lot of trouble solving abstract reasoning problems that humans are naturally fantastic at, but oh well. In recent times we have seen how the ARC-AGI 2 test, which seemed like a challenge for AI models, has become increasingly acceptable for said models. GPT-5.4 gives a new bite to that reality, and in its Pro version it already manages to solve 83.3% of the tasks (73.3% in the standard model) when in GPT-2 the rate was 52.9%. It is a simply brutal jump, and although in other tasks that jump is not so notable (it programs somewhat better according to SWE-Bench Pro, but not much), it is clear that we are facing an extraordinary model. Perfect for OpenClaw? That ability seems to come to him that was not even painted OpenClawthe AI ​​agent that has become a phenomenon in this area in recent weeks. OpenAI ended up signing its creator and is in some way the “owner” of the projectand this performance in agentic tasks is expected to be very useful for everything OpenClaw does, which is basically that: manage your machine for you. That’s where GPT-5.4 can really come into its own. And you can trust him more. According to those responsible for OpenAI, GPT-5.4 is now better at answering questions that require seeking information from multiple sources, and “identifying the most relevant ones, particularly for “needle in a haystack” type questions, and synthesizing them into a clear and well-reasoned answer.” What’s more: they rate it as the model most focused on answering based on facts and say that it is 33% less likely to answer something that is false compared to GPT-5.2. But be careful: it is very, very expensive. These capabilities, however, will not come cheap. With this launch OpenAI has updated its prices, and it has done so by making it clear that if you want the best, you will have to pay for it. The “standard” GPT-5.4 model costs $2.50 per million input tokens and $15 for output tokens, while the Pro costs a whopping $30 and $180 respectively. Claude Opus 4.6, which was until now considered the best AI model, costs $10 per million input tokens and $25 per million output tokens: it was already expensive, but GPT-5.4 Pro leaves it almost as a “bargain” AI model. Trying to stop the bleeding. The model appears at a delicate moment. According to various sources, ChatGPT has lost 1.5 million users since announcing that they had reached an agreement with the Department of Defense. That decision provoked much criticism, a movement on networks that spoke of “cancel ChatGPT” and internal tensions. Before the scandal there was already talk of the potential appearance of GPT-5.4, but it is clear that the launch now takes on a double meaning. It doesn’t just have to be better than everyone else: it has to redeem OpenAI. And above all he needs a victory. Public perception seems clear: OpenAI has been suffering lately, whether from internal dramas, talent drains, or temporarily falling behind in the performance of its models. GPT-5.4 is not a simple evolution of its founding model, because what OpenAI needs is for this model to succeed and convince people to “love again” (figuratively, you know what we mean) ChatGPT. We’ll see if he succeeds. In Xataka | Sam Altman says he’s terrified of a world where AI companies believe themselves to be more powerful than the government. It’s just what you’re building

MásMóvil has said goodbye to triple coverage. Although it may seem like it, it is not bad news.

by

Until now, the MásMóvil group had an important competitive advantage over the rest of the MVNOs: triple coverage. The group had access to Movistar coverage when the MásOrange network was not available, a feature that the operator’s customers will no longer enjoy. But it’s not all bad news. What did we have?. In April 2024, Orange and MásMóvil merged, creating MoreOrangebecoming the first operator in Spain in number of clients. From that moment on, the group’s clients began to enjoy triple coverage: The Orange one That of the extinct Yoigo And that of Movistar Automatically and depending on network availability, our SIM card was capable of switching between the three connectivity modules. what has happened. As explained Xataka Mobilewith statements from MasOrange, the Movistar network will no longer be available on the group’s brands. “In the old MásMóvil brands that are under the MasOrange umbrella, we are stopping using Movistar’s 4G network because our customers already have complete coverage through the MasOrange network (4G, 5G SA, 5G Advanced, depending on each case), our own, modern and high-performance network. Our commitment is that all customers always have the best possible experience.” Why is it important. Although the loss of access to the Movistar network may seem like a step backwards, the move is logical. The group has been strengthening its network coverage for years, announcing just a few hours ago important advances in its first 5G Advanced network. MásOrange network coverage map. Behind this name is the prelude to the arrival of the future 6G generation and, currently, MasOrange allows access to speeds greater than 5 Gbps with latencies of less than a second. In other words: they let Movistar go to make room for their best network. An aggressive strategy. During the last year, MásMóvil has centralized efforts to be the most competitive MasOrange low-cost. Focusing on this OMV, saying goodbye to historical ones like Lycamobileand with key alliances like the one they have with Starlink to provide its clients with satellite connectivity in the future, this goodbye is not a loss for the group: it is a strategy. In Xataka | The best cheap mobile contract rates in 2026 and comparator with all telephone companies

US sanctions are collapsing China’s factories. It’s bad news for the rest of the world

by

The US has intensified in recent years its tariff policy against China. Under the shield of “national security reasons,” the Trump administration has attempted to isolate China from essential components to create cutting-edge technology. The play didn’t go too welland China is at its best moment of national production. So much so that the capacity of its factories is reaching the limit. There are those who warned. Lip-Bu Tan, CEO of Intel, warned at the beginning of February in his statements. He pointed out that the US blockade is only achieving the opposite effect, driving giants like Huawei to develop silently and accelerating the race for China to obtain the capacity to make three nanometer chips. SMIC confirmed it. He SMIC report corresponding to the fourth quarter of 2025 is a perfect summary of China’s efforts to one day end up leading the semiconductor race. China doesn’t just want to make chips for mobile phones: it wants to dominate the semiconductors that support AI, cars, telecommunications, industry, energy and defense: because whoever controls these chips controls technological power. The key data. That SMIC’s profits have grown by 39% in the last year is quite revealing, but that the capacity of its factories has risen to 93.5% is even more so. In other words, the Chinese company is practically at the limit of its production capacity, having to satisfy the demanding demands of both the government and local companies. How does this affect me?. Among the key sectors that China wants to lead is AI. And this one needs many, many chips. So much so that SMIC has warned that the demand for them is being so enormous that the rest of the consumer electronics orders are being compromised. This ends up translating into delays in supply, price increases and something that we have been warning about for months: basic components such as RAM, SSD memories and so on. They are going to be more expensive than ever. Without help from anyone. China, without access to ASML’s most advanced machines, is achieving alternative routes for your manufacturing processes. Although some of its manufacturers are still in collaboration with giants like TSMC (case of Xiaomi with “its” XRing 01 chip, manufactured by TSCM in 3nm), the plan is to be completely self-sufficient. Something that they will end up achieving, sooner or later. In Xataka |

Great white sharks are appearing off the Alicante coast. The problem is that we don’t know if it’s good news or bad news.

by

On April 20, 2023, by pure chance, some fishermen caught a juvenile-sized white shark. No one would have been surprised if it weren’t for the fact that the fishermen were in Spanish waters, right in front of the Alicante Cape of La Nao. Two meters 10 centimeters of white shark in the middle of the Mediterranean, what was happening here? Do we have to worry? That is the question that was asked at the Spanish Institute of Oceanography and, in collaboration with the University of Cádiz, has carried out a deep review of the presence of white sharks in the Mediterranean Sea. It is not something superficial: they have collected all the records (direct and indirect) from 1862 to 2023 and have reached a surprising conclusion. The presence of this type of specimen has been “persistent” (although “extremely rare”) in the Spanish Mediterranean. It is not something, a priori, worrying. As explained José Carlos Báezresearcher at the IEO-CSIC, “we have only found two attacks: one in 1862, in which a person died in Malaga who was swimming, and another in the eighties, when a shark bit a surfer’s board in Tarifa and caused serious injuries.” But the problem is not that. And, although “with the available data, it is not possible to affirm that the Mediterranean white shark population is recovering”, it is inevitable to think about what will happen in an increasingly warmer sea. In the end, “the presence of young individuals provides key information about the demographic structure of the species” and, one way or anotherthis leads us to seriously consider the risks of having breeding spaces in Spanish waters. However, everything seems to indicate that there is a relationship between the presence of the shark and the routes of the bluefin tuna. If so, it would be another symptom of the problems that sharks have to keep their populations healthy and robust. Should we worry? It doesn’t seem like it. Against the media angle about the “return of the monster”, international evidence tells us that attacks are extremely rare and the role of sharks in the conservation of aquatic ecosystems is very important. Be that as it may, monitoring and conservation programs must be developed. And it has to be done soon. Image | Oleksandr Sushko In Xataka | The white shark is an exceptional swimmer. Its secret is in its “teeth”

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.